Abstract
268
Objectives: Phosphodiesterase-4 (PDE4) has emerged as a therapeutic target for central nervous system, inflammatory, and respiratory diseases. PDE4 has 4 isoforms (A-D) among which B and D are predominant in brain. Allosteric modulators for PDE4D are sought as candidate cognition enhancers and antidepressants with low potential for emesis side-effect [1]. [11C](R)-Rolipram readily images PDE4 in primate brain [2, 3]. However, [11C](R)-Rolipram is a non-subtype selective inhibitor and has no utility for showing target engagement by experimental drugs at subtypes of PDE4. A radioligand selective for imaging PDE4D with positron emission tomography (PET) would be useful for drug development and for studying PDE4D regulation in neuropsychiatric disorders. Here we describe the synthesis and evaluation of candidate radioligands ([11C]1−6 and [18F]7 for PET imaging of PDE4D in primate brain.
Methods: Ligands 1−7, des-methyl precursors for labeling with carbon-11 (8−13), and an iodonium salt for fluorine-18 labeling (14) were synthesized de novo. 11C-Labeled tracers were prepared by treating a desmethyl precursor with [11C]iodomethane in the presence of a base at RT (e.g., [11C]6 and 7 in Figure 1A] or at 80 °C for 5 min. For 18F-labeling, iodonium salt precursor plus TEMPO in DMSO was added to dry[18F]fluoride ion (~ 8 GBq) and heated with microwaves (130 ⁰C for 2.5 min) [Figure 1B]. Rhesus monkeys were administered with radioligand (0.15−0.3 GBq i.v.) and PET scans were acquired on a Focus 220 scanner. The monkey was administered either rolipram (0.1, 0.2, 1 mg/kg, i.v.) or BPN14770, a selective PDE4D negative allosteric modulator (3 mg/kg, i.v.). Concurrent arterial blood sampling was performed to obtain the metabolite-corrected input function for quantification. Total distribution volumes (VT) for different regions were calculated using a two-tissue compartmental model.
Results: In vitro enzyme inhibition assays revealed that 1igands 1-7 are potent (PDE4D IC50 values range subnanomolar to low nanomolar) and selective to PDE4D over PDE4B. [11C]1-6 were obtained in 10-20% decay-corrected formulated yields in ~ 40 min synthesis time with a molar activity (Am) of > 200 GBq/µmol. The yield of [18F]7 was 10% with Am of ~ 98 GBq/µmol at EOS (~ 90 min synthesis time). At baseline, each radioligand readily entered monkey brain with quick washout. However, receptor pre-block revealed no appreciable specific signal for [11C]1-[11C]4 and [18F]7. Radioligands [11C]5 and [11C]6 performed best giving moderate brain uptake (SUV ~3) and appreciable specific signal [Figure 2]. [11C]6 uptake was also appreciably blocked by BPN14770 [Figure 2C]. VT time stability of [11C]5 and [11C]6 indicated possible accumulation of radiometabolites in the brain but was adequate for quantification as the normalized VT improves towards the end of the scan (120 min). Conclusions: [11C]5 and [11C]6 are the best among the radioligands evaluated with [11C]6 being the slightly superior PET radioligand and may be used for occupancy and target engagement by experimental drugs at PDE4D in monkey.Acknowledgements: The Intramural Research Program of NIH (NIMH).References [1] Burgin AB et al., 2010 Nat. Biotechnol. 28, 63. [2] Fujita M et al., 2012 Biol. Psychiatry, 72, 548. [3] Ooms M et al., 2019 J Cerebr Blood Flow Metab., 39, 1306.